Difference between revisions of "Alternatively, making the initial path with PATHSAMPLE itself (CHARMM)"

From Docswiki
Jump to navigation Jump to search
import>Jmc49
(Created page with "To run PATHSAMPLE with CHARMM to set up a database from scratch, you need to have the following files in a directory together: * '''input.crd''' - as usual * '''perm.allow''…")
 
import>Jmc49
m
 
(4 intermediate revisions by the same user not shown)
Line 1: Line 1:
To run [[PATHSAMPLE]] with CHARMM to set up a database from scratch, you need to have the following files in a directory together:
+
To run [[PATHSAMPLE]] with CHARMM to set up a database from scratch, and make the initial connection, you need to have the following files in a directory together:
   
 
* '''input.crd''' - as usual
 
* '''input.crd''' - as usual
 
* '''perm.allow''' - as usual
 
* '''perm.allow''' - as usual
  +
* '''odata.start''': as usual for local minimization, but with DUMPDATA and ENDHESS added to produce the relevant data (possibly with ENDNUMHESS too if you're using numerical second derivatives)
* '''odata.start''':
 
 
<pre>
 
<pre>
 
DUMPDATA
 
DUMPDATA
Line 15: Line 15:
 
MAXERISE 1.0D-4 1.0D0
 
MAXERISE 1.0D-4 1.0D0
 
STEPS 400
 
STEPS 400
CHARMMTYPE top_all22_prot.inp par_all22_prot.inp
+
CHARMMTYPE toph19_eef1_perm.inp param19_eef1_perm.inp
 
CHARMM
 
CHARMM
... the usual charmm stuff
+
... the usual charmm stuff, but making sure that start.crd is the file from which the coordinates are read in, for example:
  +
OPEN UNIT 20 NAME start.crd READ CARD
  +
...
 
</pre>
 
</pre>
 
* '''start.crd''' - the desired first endpoint, in CHARMM's .crd format.
 
* '''start.crd''' - the desired first endpoint, in CHARMM's .crd format.
* '''odata.finish''':
+
* '''odata.finish''': as for odata.start
 
<pre>
 
<pre>
 
DUMPDATA
 
DUMPDATA
Line 32: Line 34:
 
MAXERISE 1.0D-4 1.0D0
 
MAXERISE 1.0D-4 1.0D0
 
STEPS 400
 
STEPS 400
CHARMMTYPE top_all22_prot.inp par_all22_prot.inp
+
CHARMMTYPE toph19_eef1_perm.inp param19_eef1_perm.inp
 
CHARMM
 
CHARMM
... the usual charmm stuff
+
... the usual charmm stuff, but making sure that finish.crd is the file from which the coordinates are read in, for example:
  +
OPEN UNIT 20 NAME finish.crd READ CARD
  +
...
 
</pre>
 
</pre>
 
* '''finish.crd''' - the second endpoint in .crd format.
 
* '''finish.crd''' - the second endpoint in .crd format.
* '''odata.connect''' - specified parameters for minima connection attempts in [[OPTIM]]; you could just copy the odata file from [[Finding an initial path with OPTIM and starting up PATHSAMPLE]] to odata.connect in your working directory.
+
* '''odata.connect''' - specified parameters for minima connection attempts in [[OPTIM]]; you could just copy the odata file from [[Finding an initial path with OPTIM and starting up PATHSAMPLE (CHARMM)]] to odata.connect in your working directory.
 
* '''pathdata''' - the [[PATHSAMPLE]] input file (see documentation):
 
* '''pathdata''' - the [[PATHSAMPLE]] input file (see documentation):
 
<pre>
 
<pre>
Line 52: Line 56:
 
TEMPERATURE 0.592
 
TEMPERATURE 0.592
 
COMMENT MAXTSENERGY -80.0
 
COMMENT MAXTSENERGY -80.0
JOBSPERNODE 1
+
COMMENT JOBSPERNODE 1
  +
PBS
 
CONNECTIONS 1
 
CONNECTIONS 1
 
COPYFILES input.crd perm.allow
 
COPYFILES input.crd perm.allow
Line 63: Line 68:
 
The DIJINITSTART keyword is the one that sets off OPTIM local minimizations for start.crd and finish.crd, takes those minima as the two end points, then tries iteratively to make a connected path between them by setting up and running some [[OPTIM]] connection-making jobs between likely pairs of minima. It analyses the output from those jobs and adds all the min-TS-min triples found to the growing database. The run will stop when you have made a fully connected discrete path between the end points.
 
The DIJINITSTART keyword is the one that sets off OPTIM local minimizations for start.crd and finish.crd, takes those minima as the two end points, then tries iteratively to make a connected path between them by setting up and running some [[OPTIM]] connection-making jobs between likely pairs of minima. It analyses the output from those jobs and adds all the min-TS-min triples found to the growing database. The run will stop when you have made a fully connected discrete path between the end points.
   
The COPYOPTIM keyword is useful for debugging to see how well the connection-making attempts are doing. Have a look at the OPTIM.connect.* files, as described in [[Finding an initial path with OPTIM and starting up PATHSAMPLE]]. It is not a keyword that you want to leave in pathdata when doing the later sampling part of the calculations though as you'll get loads of random files cluttering up your working directory!!!
+
The COPYOPTIM keyword is useful for debugging to see how well the connection-making attempts are doing. Have a look at the OPTIM.connect.* files, as described in [[Finding an initial path with OPTIM and starting up PATHSAMPLE (CHARMM)]]. It is not a keyword that you want to leave in pathdata when doing the later sampling part of the calculations though as you'll get loads of random files cluttering up your working directory!!!
   
 
Also, make sure you set the full path of your COPTIM executable correctly on the EXEC line.
 
Also, make sure you set the full path of your COPTIM executable correctly on the EXEC line.

Latest revision as of 18:59, 1 February 2016

To run PATHSAMPLE with CHARMM to set up a database from scratch, and make the initial connection, you need to have the following files in a directory together:

  • input.crd - as usual
  • perm.allow - as usual
  • odata.start: as usual for local minimization, but with DUMPDATA and ENDHESS added to produce the relevant data (possibly with ENDNUMHESS too if you're using numerical second derivatives)
DUMPDATA
ENDHESS
UPDATES 10000
BFGSCONV 1.0D-6
BFGSMIN 1.0D-6
CONVERGE 1.0D-6
BFGSSTEPS 60000
DEBUG
MAXERISE 1.0D-4 1.0D0
STEPS 400
CHARMMTYPE toph19_eef1_perm.inp param19_eef1_perm.inp
CHARMM
... the usual charmm stuff, but making sure that start.crd is the file from which the coordinates are read in, for example:
OPEN UNIT 20 NAME start.crd READ CARD
...
  • start.crd - the desired first endpoint, in CHARMM's .crd format.
  • odata.finish: as for odata.start
DUMPDATA
ENDHESS
UPDATES 10000
BFGSCONV 1.0D-6
BFGSMIN 1.0D-6
CONVERGE 1.0D-6
BFGSSTEPS 60000
DEBUG
MAXERISE 1.0D-4 1.0D0
STEPS 400
CHARMMTYPE toph19_eef1_perm.inp param19_eef1_perm.inp
CHARMM
... the usual charmm stuff, but making sure that finish.crd is the file from which the coordinates are read in, for example:
OPEN UNIT 20 NAME finish.crd READ CARD
...
DIJINITSTART exp
COMMENT WARNING: DIJINITSTART runs overwrite all output, if in doubt, use DIJINITCONT 
COMMENT DIJINITCONT exp
CYCLES 1000
PAIRLIST 1
NATOMS         33
ETOL           1.0D-4
ITOL 1.0D1
GEOMDIFFTOL 1.0D-1
PLANCK 9.546E-14
TEMPERATURE   0.592
COMMENT MAXTSENERGY -80.0
COMMENT JOBSPERNODE    1
PBS
CONNECTIONS    1
COPYFILES input.crd perm.allow 
COPYOPTIM
EXEC /home/csw34/bin/C35OPTIM
PERMDIST
CHARMM

The DIJINITSTART keyword is the one that sets off OPTIM local minimizations for start.crd and finish.crd, takes those minima as the two end points, then tries iteratively to make a connected path between them by setting up and running some OPTIM connection-making jobs between likely pairs of minima. It analyses the output from those jobs and adds all the min-TS-min triples found to the growing database. The run will stop when you have made a fully connected discrete path between the end points.

The COPYOPTIM keyword is useful for debugging to see how well the connection-making attempts are doing. Have a look at the OPTIM.connect.* files, as described in Finding an initial path with OPTIM and starting up PATHSAMPLE (CHARMM). It is not a keyword that you want to leave in pathdata when doing the later sampling part of the calculations though as you'll get loads of random files cluttering up your working directory!!!

Also, make sure you set the full path of your COPTIM executable correctly on the EXEC line.